Return to index: [Subject] [Thread] [Date] [Author]

RE: Complexity of Code

[Subject Prev][Subject Next][Thread Prev][Thread Next]
While I think you are coming from the right direction, I would have to
disagree with you when you indicated the code[s] "are made to be optimal for
computer programming." This can't be farther from the truth. As one with
extensive experience in spreadsheet programming, and as the co-author of
Multi-Lat - the freeware spreadsheet that is intended to provide full
compliance design for a multi-story wood framed structure - I discovered
very quickly that many of the code formulas are derivations of basic
mechanics of materials, but which have been derived so many times that the
original meaning was lost. The perfect example was our List discussion some
months ago on the discussion of diaphragm deflection calculations. In
particular the multi-part deflection formula. The first part of the formula
looks that the chord deflection. The problem with the way it is written is
that in weeks of discussing the formula, only one person was able to track
it back to the basic 5wl^4/(384EI). The majority of engineers who responded
believed that the formula was based on empirically tested data that led to
the creation of constants.
The code was not clear on this and none of the ICBO Design guides attempted
to rebuild the formula so that engineers could backtrack and understand how
the derivation occurs. Yet one engineer did and when he provided the
evidence, it was easy to establish the formula in programs like Mathcad or
Spreadsheets and obtain accurate results.

I don't agree that the code simplifies the modeling process. In fact, that
is the weakness of the code. If more thought and attention was given to
rebuilding the formula's so that many hours of research was not required to
find this simple relationship, the majority of engineers would be on much
firmer ground. Understanding this creates a tremendous shortcut to
developing a professional intuition for how the materials work and what kind
of performance is expected.

I don't disagree with you that an engineer should never place his or her
faith in a computer program without developing their own instant for what
the formulas relationship to practical applications should be. This is an
important step.

On the other hand, Hewlett Packard and Texas Instruments are promoting
calculators as learning tools for the middle school and high school
students. They have developed education programs to encourage teachers to
use their products to teach students rather than fearing that the tool will
be misused. The tool is only as good as the mind that controls it.

I've learned the provisions of the new code by recreating the flow charts
necessary to develop the path of logic to go from one step to the next.
Specifically, I had to not only recreate the basic multi-story lateral load
distribution, but I also had to work out the algorithms that would
distribute force by tributary area in three dimensions (at each level and in
multi-level structures) - especially when walls can be balanced by adding
and removing sections.

There is no right or wrong answer to the use of computer tools, however, I
do firmly believe that you are incorrect when you dig deeper into the code.
It is not set up to develop a logic model which can be converted to a usable

Dennis S. Wish, PE

-----Original Message-----
From: Craig J Slama [mailto:cslama(--nospam--at)]
Sent: Tuesday, February 06, 2001 7:51 PM
To: seaint(--nospam--at)
Subject: Re: Complexity of Code

Thanks for posting this letter.  I think that it addresses a very
critical issue.
Being a very young engineer, I should have a high regard for all things
that can
be modelled with a computer.  My previous boss, however, taught me the
errors that can be "hidden" from designers by using a computer model.

The formulas in the new code are made to be optimal for computer
but are so full of greek letters that subscripts are needed on both the
greek and
english alphabet.  I am just waiting for a factor such as Phi subscript
Presumably, modern computers are far better able to understand ancient
greek than
they can understand english.

In my current designs I am often questioned about the acceptability of
hand calculations.
However, my old friend M=wl^2/8 and PL/4, coupled with Stress=force/area
helped me in all but a few cases.  I have never gotten the relative
magnitudes of reactions
wrong, nor have I undersized beam/span ratios.  I am not saying that I
have not made
mistakes, but instead, that the mistakes I made were easily visible, and
obvious from
the calculations.

Currently, I am doing a lot of inter-company peer review, and I am
finding that the various
greek letters are doing a wonderful job of disguising many mundane
mistakes, to the point
where the designers become so engrossed with either the minor factors or
the computer
model, that they lose sight of far greater issues.

I would definately prefer to see all current graduates not being taught
modelling, rather,
being taught simple statics and structural analysis.  Over 98% ( OK, i
never actually
measured this number) of my designs have been done with a calculater,
sketch pad,
and a set of tables.

It is unfortuante that codes complicate issues overmuch, and that these
can be easily glossed over in a computer; it is also unfortunate that I
have rambled on
for so long.

Craig Slama, PEng

Lynn wrote:

> Dear list members:
> Peter Ehlen, A Structural Engineer here in Santa
> Barbara, my former boss (many years ago now), wrote a
> letter recently to SEAOC in response to the 2000 Needs
> Assessment Survey that was sent out to all SEAOC
> members.  I recently was given the opportunity to read
> this letter and felt it was important to share it with
> the list members.
> Peter Ehlen has given me permission to post this letter
> to the list.  I am in complete agreement with Peter on
> this issue of Code complexity.  The letter is as
> follows:
> Gentlemen:
> I appreciate this assessment effort and welcome the
> opportunity to share with you some thoughts that I have
> regarding what our associates can do in the future.  My
> concern lies with the effectiveness of the structural
> design portions of the building code and their general
> drift toward a much more complicated document that is
> more and more difficult and time consuming to
> understand and apply.  A simple comparison of the
> thickness of the 1955 Uniform Building Code, that I
> used when I graduated from college, and the thickness
> of Volumes 1, 2, and 3 of the current code give simple
> evidence of the additional level of complexity that
> exists today.
> I believe that those who write building codes may be
> losing sight of why the building code exists in the
> first place.  The code's purpose is to create safe
> buildings for people to live and work in.  In order for
> this to happen, the code must be easy to understand by
> a wide cross section of design skill levels.  Instead,
> the code now has become a kind of showplace for the
> most sophisticated design procedures and a battleground
> between material manufacturers who want their products
> to be used in the construction of buildings.  It is my
> experience that fewer and fewer people involved in the
> structural design of buildings have a clear
> understanding of the meaning of the code.  This
> included design engineers and plan check engineers.
> More and more time is required simply to try to
> understand the code.  Evidence of the increased
> complexity is the fact that for many sections of the
> code it is found necessary to write design aid
> documents and commentaries that are contained in the
> same document as the code.  Also, there is a general
> drift toward relying more and more on material
> manufacturer documents such as the National Design
> Standards for detailed design requirements.  It is
> really becoming a full-time job simply to find where
> things are in the code and to understand their
> meaning.  I am sure that members of the Structural
> Engineers Association of California who are active in
> the building design field are aware of the time and
> complexity associated with the recent codes and,
> ironically, the difficulty the codes present in
> fulfilling the basic purpose of the code.
> One of the problems with the more complicated and
> time-consuming codes is that they take time away from
> other portions of the structural engineering practice
> that are of equal importance.  These include having the
> time for design engineers to really think how a
> building is behaving and not be completely absorbed in
> the literature of the code.  In addition, less and less
> time is being spent by structural engineering design
> firms in providing construction observation, which is
> essential to the realization of a safe building that
> will successfully perform in the years ahead.
> I wish it to be understood that I have the highest
> respect for all the research that is being done in all
> areas of structural analysis and design.  My concern is
> that these extensive design procedures should not be a
> part of the code but should be used as a basis for
> insuring that more simplified procedures contained in
> the building code are in fact conservative.  Any
> individual who has a special design problem or has a
> particularly significant structure, there is always the
> opportunity to rely on this more advanced state of
> knowledge as they pursue the structural analysis and
> design of these buildings.  However, most buildings
> designed and built in California are less than three
> (3) stories and of fairly simple structural assembly.
> A much less rigorous code easily comprehensible by a
> wide cross section of structural designers is needed.
> It is my understanding that those who participate in
> the writing of structural codes consists of people in
> the academic field, those in the material supply
> fields, and I suspect a relatively small number of
> practicing engineers.  This disparity is due in part to
> the fact that the academic people and material supply
> people are financed by those respective institutions
> whereas the practicing structural engineer, who has
> excellent knowledge about how the code should be
> written, consists of small cadre of admirable
> individuals who donate a good portion of the their time
> or whose companies underwrite a portion of their time.
> In response to your Needs Assessment Survey, I strongly
> urge the Structural Engineers Association of California
> to take an active role in the creation of a building
> code, which is easily comprehensible by a wide cross
> section of design engineers and building department
> officials.  It should be a code that allows time for
> the design engineer to think through the design process
> rather than simply comprehend reams of detailed code
> requirements.  Since we are in a visionary state with
> this needs assessment survey, it might be appropriate
> to consider financially underwriting the talents of a
> few practicing structural engineers to aggressively
> pursue the creation of a user friendly deign code that
> will have the best chance to realizing the basic
> purpose of the building code, which is to create safe
> buildings.
> Peter W. Ehlen